Abstract

Chandelier cells represent a unique type of cortical GABAergic interneuron whose axon terminals (Ch-terminals) form synapses exclusively with the axon initial segments of pyramidal cells. In this study, we have used immunocytochemistry for the high-affinity plasma membrane transporter-1 (GAT-1) to analyze the distribution and density of Ch-terminals in various cytoarchitectonic and functional areas of the human neocortex. The lowest density of GAT-1–immuoreactive (-ir) Ch-terminals was detected in the primary and secondary visual (areas 17 and 18) and in the somatosensory areas (areas 3b and 1). In contrast, an intermediate density was observed in the motor area 4 and the associative frontolateral areas 45 and 46, whereas the associative frontolateral areas 9 and 10, frontal orbitary areas 11, 12, 13, 14, and 47, associative temporal areas 20, 21, 22, and 38, and cingulate areas 24 and 32 displayed the highest density of GAT-1-ir Ch-terminals. Despite these differences, the laminar distribution of GAT-1-ir Ch-terminals was similar in most cortical areas. Hence, the highest density of this transporter was observed in layer II, followed by layers III, V, VI, and IV. In most cortical areas, the density of GAT-1-ir Ch-terminals was positively correlated with the neuronal density, although a negative correlation was detected in layer III across all cortical areas. These results indicate that there are substantial differences in the distribution and density of GAT-1-ir Ch-terminals between areas and layers of the human neocortex. These differences might be related to the different functional attributes of the cortical regions examined.

Introduction

A large variety of gamma-amino butyric acid (GABA)-ergic neurons with different morphological, molecular, and physiological characteristics has been described in the mammalian cerebral cortex. Chandelier cells are one of the best characterized cortical GABAergic interneurons because they are readily distinguished by their vertical rows of boutons in the terminal portion of their axons (Ch-terminals), which resemble candlesticks (Szentagothai and Arbib 1974; Jones 1975). Many interneurons establish synapses at different postsynaptic regions, both with pyramidal cells and interneurons. However, chandelier cells are unique because Ch-terminals only establish synapses with the axon initial segments (AIS) of pyramidal cells. Indeed, these cells represent the main source of the AIS synapses (Somogyi 1977; Fairen and Valverde 1980; Peters et al. 1982; Somogyi et al. 1982, 1985; Freund et al. 1983; DeFelipe et al. 1985, 1989; Williams and Lacaille 1992; Buhl et al. 1994; DeFelipe 1999). The AIS is a critical region in controlling cell excitability and the generation of axon potentials, thereby determining the axonal output of principal cells (Stuart and Sakmann 1994; Colbert and Johnston 1996). Thus, in contrast to interneurons that target membrane compartments of dendrites and somata, chandelier cells have traditionally been presumed to exert a strong influence on the output of pyramidal cells (Miles et al. 1996; DeFelipe 1999). Ch-terminals are found in different cortical areas and species, including rats (Somogyi 1977; Minelli et al. 1995; Tamas and Szabadics 2004), guinea pigs (Gulyas et al. 1993), cats (Fairen and Valverde 1980; Fariñas and DeFelipe 1991), rabbits (Muller-Paschinger et al. 1983), mice (Chiu et al. 2002), ferrets (Krimer and Goldman-Rakic 2001), monkeys (Somogyi et al. 1982, 1983; DeFelipe et al. 1985), and humans (Kisvarday et al. 1986; DeFelipe 1999; Lewis et al. 2005). However, it has become increasingly clear that inhibitory inputs to the AIS of pyramidal cells are not homogeneous across species, cortical regions, layers, or neuronal populations. In addition to occasional synapses from other types of interneurons (e.g., Gonchar et al. 2002), the single AIS of pyramidal cells may be innervated by one or a few chandelier cells (reviewed in DeFelipe 1999). Moreover, the number of synaptic inputs to the AIS differs as a function of age, location, and the projection target of the pyramidal cells (DeFelipe et al. 1985; Fariñas and DeFelipe 1991; Cruz et al. 2003). Chandelier cells are chemically heterogeneous and it has been shown that they express different combinations of substances in distinct layers and species, including the GABA transporter GAT-1; the calcium-binding proteins parvalbumin (PV) and calbindin D-28k; the peptide corticotrophin releasing factor; and the polysialylated form of the neural cell adhesion molecule (DeFelipe et al. 1989; Lewis et al. 1989; Lewis and Lund 1990; Schmidt et al. 1993; Conde et al. 1994; del Rio and DeFelipe 1994; Woo et al. 1998; DeFelipe 1999; Melchitzky et al. 1999; Arellano et al. 2004). Nevertheless, no systematic studies on the distribution and the density of Ch-terminals in the various cortical areas of the human neocortex have been performed.

In the present study, we have used GAT-1 immunocytochemistry to quantify the density of Ch-terminals because GAT-1–immuoreactive (-ir) Ch-terminals are easy to distinguish from other punctate terminal-like labeled structures in the neuropil (DeFelipe and Gonzalez-Albo 1998). We have examined the distribution of Ch-terminals as well as the density of GAT-1-ir Ch-terminals with respect to the total neuron density in each layer. The results reveal that there are significant differences in the distribution and density of GAT-1-ir Ch-terminals between the different areas and layers of the human neocortex.

Materials and Methods

In this study, we have used human autopsy brain tissue obtained 2–3 h post-mortem (kindly supplied by Dr R. Alcaraz, Forensic Pathology Service, Basque Institute of Legal Medicine, Bilbao, Spain), from 3 normal males who died in traffic accidents (M1, M7, and M8 aged 23, 49, and 69 years, respectively). The brain tissue was cut into 1.5-cm-thick coronal slices and initially fixed by immersion for 24 h in 4% paraformaldehyde diluted in 0.1 M pH 7.4 phosphate buffer (PB). Thereafter, small blocks of neocortical tissue were selected and postfixed for 1–2 days in the same fixative. The tissue was obtained from the somatosensory (areas 1 and 3b), visual (areas 17 and 18), motor (area 4), associative frontal (dorsolateral [areas 9, 10, 45, 46] and orbitary [areas 11, 12, 13, 14, 47]), associative temporal (areas 20, 21, 22, and 38), and limbic cingulate (areas 24, 32) neocortical areas. Blocks from the different Brodmann's areas of the cortex were selected from each brain according to their surface anatomy, using the patterns of the gyri and sulci. The identification of each cortical area was later confirmed by analyzing the distinctive cytoarchitectonic features of Nissl-stained sections.

After fixation, all the specimens were immersed in graded sucrose solutions and they were stored in a cryoprotectant solution at −20 °C. Serial sections (100 μm) of the cortical tissue were obtained using a vibratome, and the sections from each region and case were batch processed for immunocytochemical staining. The sections immediately adjacent to those stained immunocytochemically were Nissl stained in order to identify the cortical areas and the laminar boundaries.

Immunostaining

Sections were first treated for 30 min with a solution of 0.5% hydrogen peroxide and 50% ethanol in PB to inactivate the endogenous peroxidase activity. Subsequently, the sections were rinsed in PB and preincubated for 1 h at room temperature in a stock solution containing 3% normal goat or horse serum (Vector Laboratories, Burlingame, CA) in PB with Triton X-100 0.25%. Thereafter, the sections were incubated for 48 h at 4 °C in the same stock solution containing rabbit-anti-GAT-1 antiserum (1:500; Chemicon, Temecula, CA) or mouse-anti-neuron–specific nuclear protein (Neu-N) antibodies (1:4000; Chemicon). The sections were washed in PB, incubated in horse-anti-mouse or goat-anti-rabbit biotinylated secondary antibodies (1:200; Vector), and processed using the Vectastain ABC immunoperoxidase kit (Vector). Antibody labeling was visualized with 0.05% 3,3′-diaminobenzidine tetrahydrochloride (Sigma, St Louis, MO) and 0.01% hydrogen peroxide. The sections were rinsed in PB, mounted on glass slides, dehydrated, cleared with xylene, and coverslipped. Controls were included in all the immunocytochemical procedures, either by replacing the primary antibody with preimmune goat or horse serum in some sections, omitting the secondary antibody, or replacing the secondary antibody with an inappropriate secondary antibody (i.e., an antibody directed against another species). No significant immunolabeling was detected under these control conditions.

Quantification and Statistical Analysis

A BX51 Olympus microscope equipped with a motorized stage and a DP70 digital camera, and the Neurolucida package (MicroBrightField, Williston, VT) were used to plot all GAT-1-ir Ch-terminals visualized at a final magnification of ×400, in 3 randomly selected radial strips. Adjacent Neu-N–immunostained sections were used to trace the contour lines corresponding to the individual cortical layers within each strip. The density of GAT-1-ir Ch-terminals in each cortical layer was estimated by dividing the number of GAT-1-ir Ch-terminals by the surface area of each layer in the cortical strip.

With the aid of the Neurolucida software (MicroBrightField Inc., Version 6.0, Williston) and a ×4 objective, we first defined a 1-mm-wide strip across the cortex in Neu-N–immunostained sections from each cortical area and case, each of which extended from the pial surface to the white matter. We divided each strip into 1-mm-wide rectangles (counting fields), which corresponded to the radial extent of each cortical layer. The contours of these fields were then superimposed onto video images from adjacent GAT-1–immunostained sections in order to count Ch-terminals on the computer screen. The Neurolucida package permits the complete surface of the previously defined counting fields to be scanned (×40 objective), with successive and nonoverlapping frames (17 250 μm2). All Ch-terminals throughout the entire thickness of the section were counted within an unbiased counting frame (Gundersen 1977). Ch-terminals touching the exclusion lines were not counted.

The neuronal density was estimated stereologically in each cortical area and layer by counting the nuclei in 30-μm-thick randomly selected optical dissectors of Neu-N–immunostained sections using a ×100 oil-immersion objective (West and Gundersen 1990). A minimum of 30 optical dissectors per layer and cortical area were studied using a motorized stage and the Olympus Cast grid system (Denmark).

We applied a statistical factorial model that was designed with nested multiple comparisons of 4 factors, 2 fixed (layer and area) and 2 random (human and pseudorepliques), each measurement made in every layer, cortical area, and case. In this model we excluded either layer IV when examining differences between the areas or areas 4 and 24 (agranular areas) to detect differences between the layers. The SAS system 8v (SAS Institute Inc. Cary, NC) was used in this analysis. One-way analysis of variance with T3 Dunnet post hoc comparison was then performed with the SPSS statistical package software (SPSS Inc., Chicago, IL) in order to study the differences between the layers and areas. Spearman's tests were used to study the possible correlations.

To generate the figures, light microscopic images were captured using a digital camera (Olympus DP50) attached to an Olympus light microscope, and Adobe Photoshop 7.0 software was used to generate the figure plates (Adobe Systems Inc., San Jose, CA).

Results

In the human cerebral cortex, GAT-1 immunocytochemistry labels numerous terminal-like puncta in the neuropil and around unstained cell bodies throughout cortical layers I–VI (Figs 1 and 2). However, as described previously the elements that are most intensely labeled are the Ch-terminals (DeFelipe and González-Albo 1998). GAT-1-ir Ch-terminals were found in all cortical areas but by plotting their distribution it became apparent that there were dramatic differences between each cortical area and layer (Figs 3 and 4).

Figure 1.

Photomicrograph showing GAT-1-ir Ch-terminals (arrows) innervating the AIS of pyramidal cells in layer III of area 10. The asterisk indicates the position of the pyramidal cell soma. A small GAT-1-ir Ch-terminal indicated with 2 arrows also can be distinguished, despite the fact that the unlabeled cell of origin of the AIS is not visible in this section. Scale bar: 10, 5 μm.

Figure 1.

Photomicrograph showing GAT-1-ir Ch-terminals (arrows) innervating the AIS of pyramidal cells in layer III of area 10. The asterisk indicates the position of the pyramidal cell soma. A small GAT-1-ir Ch-terminal indicated with 2 arrows also can be distinguished, despite the fact that the unlabeled cell of origin of the AIS is not visible in this section. Scale bar: 10, 5 μm.

Figure 2.

Low-power photomicrographs showing the distribution of GAT-1-ir Ch-terminals (arrows) in layers II (A and B), III (C, E, and F), and VI (D) of areas 21 (A), 9(B), 10 (C), 47(D), 17 (E), and 18 (F). Note the low density of GAT-1-ir Ch-terminals in layer III of the primary (E) and secondary (F) visual areas when compared with the frontal associative areas (C and D). Scale bar: 40 μm.

Figure 2.

Low-power photomicrographs showing the distribution of GAT-1-ir Ch-terminals (arrows) in layers II (A and B), III (C, E, and F), and VI (D) of areas 21 (A), 9(B), 10 (C), 47(D), 17 (E), and 18 (F). Note the low density of GAT-1-ir Ch-terminals in layer III of the primary (E) and secondary (F) visual areas when compared with the frontal associative areas (C and D). Scale bar: 40 μm.

Figure 3.

Schematic drawings based on Neurolucida plots made with a ×40 objective showing the laminar patterns of GAT-1-ir Ch-terminals in sensory, motor, and associative frontal dorsolateral areas of the M1 brain. Note the scarcity of GAT-1-ir Ch-terminals in sensory areas. Despite the differences in density, a similar distribution of GAT-1-ir Ch-terminals was found in most areas.

Figure 3.

Schematic drawings based on Neurolucida plots made with a ×40 objective showing the laminar patterns of GAT-1-ir Ch-terminals in sensory, motor, and associative frontal dorsolateral areas of the M1 brain. Note the scarcity of GAT-1-ir Ch-terminals in sensory areas. Despite the differences in density, a similar distribution of GAT-1-ir Ch-terminals was found in most areas.

Figure 4.

Schematic drawings based on Neurolucida plots made with a ×40 objective showing the laminar patterns of GAT-1-ir Ch-terminals in associative frontal orbitary, associative temporal, and cingulate areas of the M1 brain. Note the similar distribution of GAT-1-ir Ch-terminals in the different cortical areas.

Figure 4.

Schematic drawings based on Neurolucida plots made with a ×40 objective showing the laminar patterns of GAT-1-ir Ch-terminals in associative frontal orbitary, associative temporal, and cingulate areas of the M1 brain. Note the similar distribution of GAT-1-ir Ch-terminals in the different cortical areas.

Density of GAT-1-ir Ch-Terminals

The global density of GAT-1-ir Ch-terminals was compared between each individual and it was seen to be highest in the tissue from M1 followed by that from M7. The lowest density of GAT-1-ir Ch-terminals was observed in the cortical tissue from the individual M8 (Fig. 5). However, the differences in the density of GAT-1-ir Ch-terminals were not statistically significant either between subjects or between replicates (the different measurements made in each layer, cortical area, and case). Nevertheless, there were clear differences in the mean density of GAT-1-ir Ch-terminals between different cortical areas and layers (P < 0.0001, see below). For this reason, we will present the data from both each individual case and the averages of the values obtained in the 3 cases.

Figure 5.

Chart indicating the average density ± SE of Ch-terminals immunostained for GAT-1 in the neocortex of the individuals M1, M7, and M8. Note the decline in the mean density of GAT-1-ir Ch-terminals with age, although the differences were not statistically significant.

Figure 5.

Chart indicating the average density ± SE of Ch-terminals immunostained for GAT-1 in the neocortex of the individuals M1, M7, and M8. Note the decline in the mean density of GAT-1-ir Ch-terminals with age, although the differences were not statistically significant.

When the densities observed in the 3 cases were averaged, there were clear differences in the mean density of GAT-1-ir Ch-terminals between different cortical areas (P < 0.0001, Fig. 6A). In general, the different cortical areas analyzed could be divided into 3 main groups according to the density of GAT-1-ir Ch-terminals. Group I contained the lowest density of GAT-1-ir Ch-terminals (mean ± standard error [SE]: 39.23 ± 2.84), and included the primary and secondary sensory areas, both visual areas 17 and 18 and somatosensory areas 3b and 1. In group II, the motor area 4 and associative frontolateral areas 45 and 46 displayed intermediate densities (80.15 ± 5.80). In contrast, group III displayed the highest mean density (124.66 ± 3.43) and it comprised the associative frontolateral areas 9 and 10, the frontal orbitary areas 11, 12, 13, 14, and 47, the associative temporal areas 20, 21, 22, and 38, and cingulate areas 24 and 32. No significant differences were found in the density of the GAT-1-ir Ch-terminals between individual layers or in the mean values of cortical areas within each group (Table 1). However, all the cortical areas of group I showed significant differences in the density of these terminals when compared with cortical areas in group III (Table 1). Only certain areas of group II showed significant differences in the density of GAT-1-ir Ch-terminals with the areas included in group I or III (Table 1).

Figure 6.

Charts indicating the average density ±SE of Ch-terminals immunostained for GAT-1 in the different cortical areas (A) and layers (B) of the human brains M1, M7, and M8. Note that similar values were obtained for each area in the 3 cases. When the average values from the 3 cases are considered (A), cortical areas can be grouped according to the density of GAT-1-ir Ch-terminals. Note also that similar laminar patterns of GAT-1-ir Ch-terminals were found in the 3 brains (B).

Figure 6.

Charts indicating the average density ±SE of Ch-terminals immunostained for GAT-1 in the different cortical areas (A) and layers (B) of the human brains M1, M7, and M8. Note that similar values were obtained for each area in the 3 cases. When the average values from the 3 cases are considered (A), cortical areas can be grouped according to the density of GAT-1-ir Ch-terminals. Note also that similar laminar patterns of GAT-1-ir Ch-terminals were found in the 3 brains (B).

Table 1

Comparison of the laminar density of GAT-1-ir Ch-terminals between cortical areas

graphic 
graphic 

Laminar Distribution of GAT-1-ir Ch-Terminals

When the data from all the cortical areas were averaged, each of the 3 brains showed a similar laminar distribution of GAT-1-ir Ch-terminals (Fig. 6B). The highest density of GAT-1-ir Ch-terminals corresponded to layer II (mean ± SE; 188.08 ± 7.34), which was significantly greater than that in the other layers (P < 0.0001). Layer III displayed an intermediate density of Ch-terminals (119.28 ± 5.28), significantly different to all layers except layer VI (P < 0.01). Finally, the lowest density of Ch-terminals was found in layers IV and V (80.43 ± 4.86 and 94.80 ± 4.66, respectively). Significant differences were found between the mean density of GAT-1-ir Ch-terminals in layers IV (P < 0.001) and V (P < 0.05) with that in other layers except layers V and IV, respectively. Finally, in layer VI the density of Ch-terminals (117.89 ± 5.03) was significantly different from that in all the other layers except that in layer III (P < 0.05).

The laminar distribution of GAT-1-ir Ch-terminals in sensory, motor, frontolateral associative, temporal associative, and cingulate areas is shown in Figures 3, 4, 7, and 8. Note that the distribution of these terminals is fairly similar in the 3 brains analyzed (Figs 7 and 8). The significance of the comparisons between the different layers across the different areas analyzed is illustrated in Table 1. In general, when significant differences were observed in the density of GAT-ir Ch-terminals between areas, they were mostly due to significant differences in layers III and VI. Layer II presented a homogeneously high density of GAT-1-ir Ch-terminals with no significant differences between most cortical areas. In contrast, the density of terminals in layer II of areas 1 and 18 was significantly lower when compared with most cortical areas of group III. In most areas other than the temporal lobe, a tendency toward a lower density of GAT-1-ir Ch-terminals was observed in layers IV and V, although the differences between areas did not generally reach statistical significance.

Figure 7.

Graphs indicating the density of GAT-1-ir Ch-terminals in the different layers of sensory, motor, and associative frontal dorsolateral areas of the M1, M7, and M8 brains. Note that despite the differences in density, the same laminar patterns were generally found in each cortical area in the 3 individuals.

Figure 7.

Graphs indicating the density of GAT-1-ir Ch-terminals in the different layers of sensory, motor, and associative frontal dorsolateral areas of the M1, M7, and M8 brains. Note that despite the differences in density, the same laminar patterns were generally found in each cortical area in the 3 individuals.

Figure 8.

Graphs indicating the density of GAT-1-ir Ch-terminals in the different layers of associative frontal orbitary, associative temporal, and cingulate areas in the M1, M7, and M8 brains. Note that despite the differences in density, the same laminar patterns were generally found in each cortical area in the 3 individuals. See layer I in area 38 of the M1 brain as an exception.

Figure 8.

Graphs indicating the density of GAT-1-ir Ch-terminals in the different layers of associative frontal orbitary, associative temporal, and cingulate areas in the M1, M7, and M8 brains. Note that despite the differences in density, the same laminar patterns were generally found in each cortical area in the 3 individuals. See layer I in area 38 of the M1 brain as an exception.

Correlation between the Density of GAT-1-ir Ch-Terminals and that of Neu-N-ir Neurons

In an attempt to detect possible correlations between the density of Neu-N-ir neurons and the number of GAT-ir Ch-terminals, we compared these parameters in 6 representative cortical areas: the sensory areas 17 and 18, motor area 4, associative areas 9 and 21, and cingulate area 24 (Figs 7–9). When these variables were analyzed in individual cortical layers from distinct cytoarchitectonic areas, an inverse correlation was found in layer III (Spearman's rho −0.664, P = 0.03). Thus, in areas with a higher neuronal density in layer III (i.e., areas 17 and 18) there was a lower density of GAT-1-ir Ch-terminals (Figs 7–9). Likewise, a similar trend was found in layers II and IV but this did not reach statistical significance (Spearman's rho −0.457 P = 0.56 and Spearman's rho −0.493 P = 0.62, respectively). In contrast, the number of GAT-ir Ch-terminals was not correlated with neuronal density in layers V and VI.

Figure 9.

Bar diagrams illustrating the neuronal density (M1, M7, and M8) in sensory (areas 17 and 18), motor (area 4), dorsolateral associative (area 9), temporal associative (area 21), and cingulate (area 24) areas.

Figure 9.

Bar diagrams illustrating the neuronal density (M1, M7, and M8) in sensory (areas 17 and 18), motor (area 4), dorsolateral associative (area 9), temporal associative (area 21), and cingulate (area 24) areas.

When neuronal density and the density of GAT-ir Ch-terminals were analyzed in the cortical layers of each individual cytoarchitectonic area, a direct correlation was found in areas 4, 9, 17, and 24. As such, the higher density of GAT-1-ir Ch-terminals was associated with a higher density of neurons in the layers of these areas (Spearman's rho for areas 4, 9, 17, 24 = 0.83**, 0.65**, 0.68**, 0.53*; *P < 0.05, **P < 0.01).

Discussion

Through studying GAT-1 immunohistochemistry, we have been able to quantify the density and distribution of Ch-terminals in different areas and layers of the human neocortex. This analysis has revealed both similarities and differences between distinct cytoarchitectonic areas and cortical layers. Thus, we found that the distribution of Ch-terminals is not homogeneous in the cortex and we consider that these differences are probably related to the different functional attributes of the cortical region or layer examined.

Variation in the Density of Ch-Terminals between Individuals

In this study, we found that the global density of GAT-1-ir Ch-terminals decreased with age. Indeed, there was a 15% decrease in the 69-year-old individual when compared with the tissue obtained from the 23 year old. However, this difference was not statistically significant and because we only analyzed one individual of each age, the differences could merely be due to interindividual variability. Thus, it is clear that further studies on tissue from more subjects will be necessary to verify the age-related decline in the density of GAT-1-ir Ch-terminals in the human neocortex. Nevertheless, these results are consistent with the age-related decline in GAT-1-ir Ch-terminals reported in layer III of the monkey cortex (Cruz et al. 2003). Such a decline does not seem to represent an age-dependent decrease in the number of neurons. Neither neuronal density (Pakkenberg et al. 2003) nor the density of PV-ir cell somata (including chandelier cells; Bu et al. 2003) have been seen to change as a function of age in the human neocortex. Thus, age-related axonal retraction by chandelier cells and/or a decrease of GAT-1 expression or antigenicity at Ch-terminals with age may well account for the current observations. However, because there are relatively few chandelier cells a loss of some these may pass unnoticed when the whole population of neurons is considered in stereological studies. It will be necessary to examine the pyramidal cell AISs by electron microscopy to better define the processes involved in this loss of GAT-1–labeled terminals.

Variation in the Density of GAT-1-ir Ch-Terminals between Areas

There were significant differences in the density of GAT-ir Ch-terminals between the distinct areas of the human neocortex analyzed, suggesting a regional specialization of the inhibitory circuits in which they are involved. In the 3 brains analyzed, the primary and secondary sensory areas (areas 17, 18, 3b, and 1) displayed the lowest density of GAT-1-ir Ch-terminals, the motor (area 4) and associative frontolateral areas 45 and 46 displayed intermediate values, whereas the frontolateral areas 9 and 10, the frontal orbitary areas 11, 12, 13, 14, and 47, the temporal areas 20, 21, 22, and 38, and the cingulate areas 24 and 32 contained the highest mean density of GAT-1-ir Ch-terminals. Although detailed studies of the density and distribution of Ch-terminals have not been performed previously in the human neocortex, these results are consistent with observations that there is generally a higher density of Ch-terminals in high-order association areas than in primary sensory areas in the monkey (Lewis et al. 1989; Akil and Lewis 1992; Conde et al. 1996; Elston and Gonzalez-Albo 2003). Furthermore, the results are consistent with those showing that double bouquet cells, another important component of cortical GABAergic circuits, are not homogeneously distributed throughout the neocortex but that they also display rather dramatic differences in their density between areas (Yáñez et al. 2005). Previous electron microscopy and tract tracing studies (Fariñas and DeFelipe 1991), as well as the analysis of Golgi-stained material (Lewis and Lund 1990) have shown that Ch-terminals are present in the cat and monkey visual cortex. However, it is difficult to obtain an idea about the distribution of Ch-terminals in electron microscope studies or in Golgi studies due to the inconsistency of the Golgi method and the relatively few AIS examined by electron microscopy. Therefore, the scarcity of GAT-1-ir Ch-terminals in the primary visual cortex observed here, as in other sensory areas, is best compared with other cortical areas examined using the same methodology. There are several possible explanations for the differences in the density of GAT-1-ir Ch-terminals between areas. Because not all pyramidal cells are necessarily innervated by Ch-terminals (e.g., see Fariñas and DeFelipe 1991), a smaller proportion of the pyramidal cells in sensory areas could be innervated by Ch-terminals than in associative areas. Alternatively, there may be less GAT-1 protein in Ch-terminals in sensory areas implying that many terminals in these regions contain quantities of protein that are below the levels of detection for the immunocytochemical methods used. Furthermore, it is possible that at least some Ch-terminals in sensory cortical areas could use other types of GABA transporters.

Laminar Distribution of GAT-ir Ch-Terminals in the Cerebral Cortex

The present results also indicate that there are differences in the density of GAT-1-ir Ch-terminals between the distinct cortical layers. The highest density of GAT-1-ir Ch-terminals was observed in layer II followed by layers III, V, VI, and IV. Previous studies have shown that there are differences in the density of synaptic terminals that contact the AIS of pyramidal cells located in different layers. For instance, in the monkey temporal and sensory-motor cortex, and in the cat visual cortex, the AIS of pyramidal neurons in layers II–III are more densely innervated by chandelier cell synapses than those in layers V and VI (Sloper and Powell 1979; DeFelipe et al. 1985; Fariñas and DeFelipe 1991; DeFelipe 1999). Because the pyramidal cells located in different layers project to different sites (Jones 1984; White 1989), the changes in the density of GAT-1-ir Ch-terminals observed between layers might be related to the laminar distribution of pyramidal cell populations projecting to particular sites.

In addition, it has been shown in the rat neocortex that postsynaptic GABAA receptors at the AIS of supragranular pyramidal cells are enriched with the α2 subunit, whereas in infragranular layers the α3 subunit predominates (Fritschy et al. 1998). Thus, it is possible that chandelier cells exert their activity through different types of postsynaptic GABAA receptors depending on the cortical layer in which they are found. Finally, a direct correlation was observed between the density of GAT-1-ir Ch-terminals and neuronal density in areas 4, 9, 17, and 24. This was not the case in areas 18 and 21 where there was no correlation between GAT-1-ir Ch-terminals and neuronal density. Hence, there is a remarkable heterogeneity in the density and distribution of GAT-1-ir Ch-terminals, suggesting that chandelier cells contribute differentially to cortical circuits (anatomically and physiologically) according to the cortical area or layer in which they are situated.

This work was supported by the Ministerio de Ciencia y Tecnología grants BFI 2003-01018 and BFU2006-03855 to A.M. and BFI 2003-02745 and BFU2006-13395 to J.F. We thank Laura Barrios (Department of Statistics, CTI, CSIC, Madrid, Spain) for her assistance in the statistical analysis). Conflict of Interest: None declared.

References

Akil
M
Lewis
DA
Postnatal development of parvalbumin immunoreactivity in axon terminals of basket and chandelier neurons in monkey neocortex
Prog Neuropsychopharmacol Biol Psychiatry
 , 
1992
, vol. 
16
 (pg. 
329
-
337
)
Arellano
JI
Munoz
A
Ballesteros-Yanez
I
Sola
RG
DeFelipe
J
Histopathology and reorganization of chandelier cells in the human epileptic sclerotic hippocampus
Brain
 , 
2004
, vol. 
127
 (pg. 
45
-
64
)
Bu
J
Sathyendra
V
Nagykery
N
Geula
C
Age-related changes in calbindin-D28k, calretinin, and parvalbumin-immunoreactive neurons in the human cerebral cortex
Exp Neurol
 , 
2003
, vol. 
182
 (pg. 
220
-
231
)
Buhl
EH
Halasy
K
Somogyi
P
Diverse sources of hippocampal unitary inhibitory postsynaptic potentials and the number of synaptic release sites
Nature
 , 
1994
, vol. 
368
 (pg. 
823
-
828
)
Chiu
CS
Jensen
K
Sokolova
I
Wang
D
Li
M
Deshpande
P
Davidson
N
Mody
I
Quick
MW
Quake
SR
, et al.  . 
Number, density, and surface/cytoplasmic distribution of GABA transporters at presynaptic structures of knock-in mice carrying GABA transporter subtype 1-green fluorescent protein fusions
J Neurosci
 , 
2002
, vol. 
22
 (pg. 
10251
-
10266
)
Colbert
CM
Johnston
D
Axonal action-potential initiation and Na+ channel densities in the soma and axon initial segment of subicular pyramidal neurons
J Neurosci
 , 
1996
, vol. 
16
 (pg. 
6676
-
6686
)
Conde
F
Lund
JS
Jacobowitz
DM
Baimbridge
KG
Lewis
DA
Local circuit neurons immunoreactive for calretinin, calbindin D-28k or parvalbumin in monkey prefrontal cortex: distribution and morphology
J Comp Neurol
 , 
1994
, vol. 
341
 (pg. 
95
-
116
)
Conde
F
Lund
JS
Lewis
DA
The hierarchical development of monkey visual cortical regions as revealed by the maturation of parvalbumin-immunoreactive neurons
Brain Res Dev Brain Res
 , 
1996
, vol. 
96
 (pg. 
261
-
276
)
Cruz
DA
Eggan
SM
Lewis
DA
Postnatal development of pre- and postsynaptic GABA markers at chandelier cell connections with pyramidal neurons in monkey prefrontal cortex
J Comp Neurol
 , 
2003
, vol. 
465
 (pg. 
385
-
400
)
DeFelipe
J
Chandelier cells and epilepsy
Brain
 , 
1999
, vol. 
122
  
(Pt 10):1807–1822
DeFelipe
J
Gonzalez-Albo
MC
Chandelier cell axons are immunoreactive for GAT-1 in the human neocortex
Neuroreport
 , 
1998
, vol. 
9
 (pg. 
467
-
470
)
DeFelipe
J
Hendry
SH
Jones
EG
Visualization of chandelier cell axons by parvalbumin immunoreactivity in monkey cerebral cortex
Proc Natl Acad Sci USA
 , 
1989
, vol. 
86
 (pg. 
2093
-
2097
)
DeFelipe
J
Hendry
SH
Jones
EG
Schmechel
D
Variability in the terminations of GABAergic chandelier cell axons on initial segments of pyramidal cell axons in the monkey sensory-motor cortex
J Comp Neurol
 , 
1985
, vol. 
231
 (pg. 
364
-
384
)
del Rio
MR
DeFelipe
J
A study of SMI 32-stained pyramidal cells, parvalbumin-immunoreactive chandelier cells, and presumptive thalamocortical axons in the human temporal neocortex
J Comp Neurol
 , 
1994
, vol. 
342
 (pg. 
389
-
408
)
Elston
GN
Gonzalez-Albo
MC
Parvalbumin-, calbindin-, and calretinin-immunoreactive neurons in the prefrontal cortex of the owl monkey (Aotus trivirgatus): a standardized quantitative comparison with sensory and motor areas
Brain Behav Evol
 , 
2003
, vol. 
62
 (pg. 
19
-
30
)
Fairen
A
Valverde
F
A specialized type of neuron in the visual cortex of cat: a Golgi and electron microscope study of chandelier cells
J Comp Neurol
 , 
1980
, vol. 
194
 (pg. 
761
-
779
)
Fariñas
I
DeFelipe
J
Patterns of synaptic input on corticocortical and corticothalamic cells in the cat visual cortex. II. The axon initial segment
J Comp Neurol
 , 
1991
, vol. 
304
 (pg. 
70
-
77
)
Freund
TF
Martin
KA
Smith
AD
Somogyi
P
Glutamate decarboxylase-immunoreactive terminals of Golgi-impregnated axoaxonic cells and of presumed basket cells in synaptic contact with pyramidal neurons of the cat's visual cortex
J Comp Neurol
 , 
1983
, vol. 
221
 (pg. 
263
-
278
)
Fritschy
JM
Weinmann
O
Wenzel
A
Benke
D
Synapse-specific localization of NMDA and GABA(A) receptor subunits revealed by antigen-retrieval immunohistochemistry
J Comp Neurol
 , 
1998
, vol. 
390
 (pg. 
194
-
210
)
Gonchar
Y
Turney
S
Price
JL
Burkhalter
A
Axo-axonic synapses formed by somatostatin-expressing GABAergic neurons in rat and monkey visual cortex
J Comp Neurol
 , 
2002
, vol. 
443
 (pg. 
1
-
14
)
Gulyas
AI
Miles
R
Hajos
N
Freund
TF
Precision and variability in postsynaptic target selection of inhibitory cells in the hippocampal CA3 region
Eur J Neurosci
 , 
1993
, vol. 
5
 (pg. 
1729
-
1751
)
Gundersen
HJG
Notes on the estimation of the numerical density of arbitrary profiles: the edge effect
J Microsc
 , 
1977
, vol. 
111
 (pg. 
219
-
223
)
Jones
EG
Varieties and distribution of non-pyramidal cells in the somatic sensory cortex of the squirrel monkey
J Comp Neurol
 , 
1975
, vol. 
160
 (pg. 
205
-
267
)
Jones
EG
Peters
A
Jones
EG
Laminar distribution of cortical efferent cells
Cellular components of the cerebral cortex
 , 
1984
New York
Plenum Press
Kisvarday
ZF
Adams
CB
Smith
AD
Synaptic connections of axo-axonic (chandelier) cells in human epileptic temporal cortex
Neuroscience
 , 
1986
, vol. 
19
 (pg. 
1179
-
1186
)
Krimer
LS
Goldman-Rakic
PS
Prefrontal microcircuits: membrane properties and excitatory input of local, medium, and wide arbor interneurons
J Neurosci
 , 
2001
, vol. 
21
 (pg. 
3788
-
3796
)
Lewis
DA
Foote
SL
Cha
CI
Corticotropin-releasing factor immunoreactivity in monkey neocortex: an immunohistochemical analysis
J Comp Neurol
 , 
1989
, vol. 
290
 (pg. 
599
-
613
)
Lewis
DA
Hashimoto
T
Volk
DW
Cortical inhibitory neurons and schizophrenia
Nat Rev Neurosci
 , 
2005
, vol. 
6
 (pg. 
312
-
324
)
Lewis
DA
Lund
JS
Heterogeneity of chandelier neurons in monkey neocortex: corticotropin-releasing factor- and parvalbumin-immunoreactive populations
J Comp Neurol
 , 
1990
, vol. 
293
 (pg. 
599
-
615
)
Melchitzky
DS
Sesack
SR
Lewis
DA
Parvalbumin-immunoreactive axon terminals in macaque monkey and human prefrontal cortex: laminar, regional, and target specificity of type I and type II synapses
J Comp Neurol
 , 
1999
, vol. 
408
 (pg. 
11
-
22
)
Miles
R
Toth
K
Gulyas
AI
Hajos
N
Freund
TF
Differences between somatic and dendritic inhibition in the hippocampus
Neuron
 , 
1996
, vol. 
16
 (pg. 
815
-
823
)
Minelli
A
Brecha
NC
Karschin
C
DeBiasi
S
Conti
F
GAT-1, a high-affinity GABA plasma membrane transporter, is localized to neurons and astroglia in the cerebral cortex
J Neurosci
 , 
1995
, vol. 
15
 (pg. 
7734
-
7746
)
Muller-Paschinger
IB
Tombol
T
Petsche
H
Chandelier neurons within the rabbits' cerebral cortex. A Golgi study
Anat Embryol (Berl)
 , 
1983
, vol. 
166
 (pg. 
149
-
154
)
Pakkenberg
B
Pelvig
D
Marner
L
Bundgaard
MJ
Gundersen
HJ
Nyengaard
JR
Regeur
L
Aging and the human neocortex
Exp Gerontol
 , 
2003
, vol. 
38
 (pg. 
95
-
99
)
Peters
A
Proskauer
CC
Ribak
CE
Chandelier cells in rat visual cortex
J Comp Neurol
 , 
1982
, vol. 
206
 (pg. 
397
-
416
)
Schmidt
S
Braak
E
Braak
H
Parvalbumin-immunoreactive structures of the adult human entorhinal and transentorhinal region
Hippocampus
 , 
1993
, vol. 
3
 (pg. 
459
-
470
)
Sloper
JJ
Powell
TP
A study of the axon initial segment and proximal axon of neurons in the primate motor and somatic sensory cortices
Philos Trans R Soc Lond B Biol Sci
 , 
1979
, vol. 
285
 (pg. 
173
-
197
)
Somogyi
P
A specific ‘axo-axonal’ interneuron in the visual cortex of the rat
Brain Res
 , 
1977
, vol. 
136
 (pg. 
345
-
350
)
Somogyi
P
Freund
TF
Cowey
A
The axo-axonic interneuron in the cerebral cortex of the rat, cat and monkey
Neuroscience
 , 
1982
, vol. 
7
 (pg. 
2577
-
2607
)
Somogyi
P
Freund
TF
Hodgson
AJ
Somogyi
J
Beroukas
D
Chubb
IW
Identified axo-axonic cells are immunoreactive for GABA in the hippocampus and visual cortex of the cat
Brain Res
 , 
1985
, vol. 
332
 (pg. 
143
-
149
)
Somogyi
P
Nunzi
MG
Gorio
A
Smith
AD
A new type of specific interneuron in the monkey hippocampus forming synapses exclusively with the axon initial segments of pyramidal cells
Brain Res
 , 
1983
, vol. 
259
 (pg. 
137
-
142
)
Stuart
GJ
Sakmann
B
Active propagation of somatic action potentials into neocortical pyramidal cell dendrites
Nature
 , 
1994
, vol. 
367
 (pg. 
69
-
72
)
Szentagothai
J
Arbib
MA
Conceptual models of neural organization
Neurosci Res Program Bull
 , 
1974
, vol. 
12
 (pg. 
305
-
510
)
Tamas
G
Szabadics
J
Summation of unitary IPSPs elicited by identified axo-axonic interneurons
Cereb Cortex
 , 
2004
, vol. 
14
 (pg. 
823
-
826
)
West
MJ
Gundersen
HJ
Unbiased stereological estimation of the number of neurons in the human hippocampus
J Comp Neurol
 , 
1990
, vol. 
296
 (pg. 
1
-
22
)
White
EL
Cortical Circuits: aynaptic organization of the cerebral cortex Structure, function and theory
1989
Boston
Birkhäuser
Williams
S
Lacaille
JC
GABAB receptor-mediated inhibitory postsynaptic potentials evoked by electrical stimulation and by glutamate stimulation of interneurons in stratum lacunosum-moleculare in hippocampal CA1 pyramidal cells in vitro
Synapse
 , 
1992
, vol. 
11
 (pg. 
249
-
258
)
Woo
TU
Whitehead
RE
Melchitzky
DS
Lewis
DA
A subclass of prefrontal gamma-aminobutyric acid axon terminals are selectively altered in schizophrenia
Proc Natl Acad Sci USA
 , 
1998
, vol. 
95
 (pg. 
5341
-
5346
)
Yáñez
IB
Muñoz
A
Contreras
J
González
J
Rodríguez-Veiga
E
DeFelipe
J
Double bouquet cell in the human cerebral cortex and a comparison with other mammals
J Comp Neurol
 , 
2005
, vol. 
486
 (pg. 
344
-
360
)